International Journal of Electrical and Computer Engineering (IJECE) Vol. 12, No. 1, February 2022, pp. 303~310 ISSN: 2088-8708, DOI: 10.11591/ijece.v12i1.pp303-310  303 Journal homepage: http://ijece.iaescore.com Deep segmentation of the liver and the hepatic tumors from abdomen tomography images Nermeen Elmenabawy 1 , Mervat El-Seddek 2 , Hossam El-Din Moustafa 1 , Ahmed Elnakib 1 1 Department of Electronics and Communications Engineering, Faculty of Engineering, Mansoura University, Mansoura, Egypt 2 Department of Communications and Electronics Engineering, Misr Higher Institute of Engineering and Technology, Mansoura, Egypt Article Info ABSTRACT Article history: Received Dec 1, 2020 Revised Aug 3, 2021 Accepted Aug 14, 2021 A pipelined framework is proposed for accurate, automated, simultaneous segmentation of the liver as well as the hepatic tumors from computed tomography (CT) images. The introduced framework composed of three pipelined levels. First, two different transfers deep convolutional neural networks (CNN) are applied to get high-level compact features of CT images. Second, a pixel-wise classifier is used to obtain two output- classified maps for each CNN model. Finally, a fusion neural network (FNN) is used to integrate the two maps. Experimentations performed on the MICCAI’2017 database of the liver tumor segmentation (LITS) challenge, result in a dice similarity coefficient (DSC) of 93.5% for the segmentation of the liver and of 74.40% for the segmentation of the lesion, using a 5-fold cross-validation scheme. Comparative results with the state-of-the-art techniques on the same data show the competing performance of the proposed framework for simultaneous liver and tumor segmentation. Keywords: Computed tomography Deep learning Liver Segmentation Tumors This is an open access article under the CC BY-SA license. Corresponding Author: Ahmed Elnakib Department of Electronics and Communications Engineering, Faculty of Engineering Mansoura University Elgomhouria St., Mansoura 35516, Egypt Email: nakib@mans.edu.eg 1. INTRODUCTION According to the World Health Organization (WHO), liver cancer is the main cause of cancer deaths among all types of cancers. Worldwide, around 800,000 cases of liver cancer are diagnosed each year, accounting for around 700,000 deaths [1]. In 2019, the American Cancer Society (ACS) estimated around 42,030 new cases for primary liver cancer and intrahepatic bile duct cancer in the United States, with around 31,780 deaths [1]. These metrics reflect the epidemic inflation of liver cancer. Computed tomography (CT) imaging is usually used for liver segmentation and/or liver cancer detection. However, manual segmentation of the liver and/or the liver tumors from CT images consumes a lot of time and suffers from observer variability. Therefore, the design of efficient computer aided diagnostic (CAD) systems, to assist the radiologists for liver segmentation and/or liver cancer segmentation, is a widely investigated open research problem. Throughout literature, different methodologies have been utilized for liver segmentation and/or for liver cancer segmentation. These methods can be categorized as traditional methods or deep learning methods. Traditional approaches usually extract features, e.g., intensity, texture, shape, from liver CT images and use a classifier based on these features to perform the segmentation process. On the other hand, deep learning methods usually use convolutional neural networks (CNN) that consist of a number of convolutional layers for extracting low-level and high-level features for the liver CT images and fully connected layers to encode a compact feature set for the segmentation process.